Fix timeout calc of nested resource groups

The timeout calculation of nested resource groups are using seconds but
then it passing the value in seconds to the stack resource that is
expected it in minutes. This leads to a rapidly increasing timeout value
in each nesting levels.

This patch converts the seconds to minutes before passing it forward.

Story: 2009237
Task: 43372
Change-Id: I3f132d4889723f7b4a3a416779ac5ee7663249b8

heat/engine/resources/openstack/heat/resource_group.py

@@ -15,6 +15,7 @@ import collections
 import copy
 import functools
 import itertools
+import math
 
 from oslo_log import log as logging
 
@@ -433,18 +434,18 @@ class ResourceGroup(stack_resource.StackResource):
                 return False
         return True
 
-    def _run_to_completion(self, template, timeout):
+    def _run_to_completion(self, template, timeout_mins):
         updater = self.update_with_template(template, {},
-                                            timeout)
+                                            timeout_mins)
 
         while not super(ResourceGroup,
                         self).check_update_complete(updater):
             yield
 
-    def _run_update(self, total_capacity, max_updates, timeout):
+    def _run_update(self, total_capacity, max_updates, timeout_mins):
         template = self._assemble_for_rolling_update(total_capacity,
                                                      max_updates)
-        return self._run_to_completion(template, timeout)
+        return self._run_to_completion(template, timeout_mins)
 
     def check_update_complete(self, checkers):
         for checker in checkers:
@@ -776,13 +777,18 @@ class ResourceGroup(stack_resource.StackResource):
 
         batches = list(self._get_batches(self.get_size(), curr_cap, batch_size,
                                          min_in_service))
-        update_timeout = self._update_timeout(len(batches), pause_sec)
+        update_timeout_secs = self._update_timeout(len(batches), pause_sec)
+
+        # NOTE(gibi) update_timeout is in seconds but the _run_update
+        # eventually calls StackResource.update_with_template that takes
+        # timeout in minutes so we need to convert here.
+        update_timeout_mins = math.ceil(update_timeout_secs / 60)
 
         def tasks():
             for index, (curr_cap, max_upd) in enumerate(batches):
                 yield scheduler.TaskRunner(self._run_update,
                                            curr_cap, max_upd,
-                                           update_timeout)
+                                           update_timeout_mins)
 
                 if index < (len(batches) - 1) and pause_sec > 0:
                     yield scheduler.TaskRunner(pause_between_batch, pause_sec)

heat_integrationtests/functional/test_resource_group.py

@@ -181,15 +181,7 @@ resources:
                 nested_stack = self.client.stacks.get(res.physical_resource_id)
                 timeouts.add(nested_stack.timeout_mins)
 
-        # FIXME(gibi): This is bug story/2009237 as the timeout calculation of
-        # nested resource groups are broken. It is using seconds for the
-        # calculation but then pass that to functions expecting timeouts in
-        # minutes leading to increasing timeout value with each nesting level.
-        self.assertEqual({35939, 599}, timeouts)
-        # After the bug is fixed we expect timeouts less than the overall
-        # timeout requested for the whole stack.
-
-        # self.assertEqual({10}, timeouts)
+        self.assertEqual({10}, timeouts)
 
     def test_update_increase_decrease_count(self):
         # create stack with resource group count 2